Exchange of extracellular domains of CCR1 and CCR5 reveals ...

795

Cardiovascular Biology and Cell Signalling

Exchange of extracellular domains of CCR1 and CCR5 reveals confined functions in CCL5-mediated cell recruitment Birgit K. Kramp1; Remco T. A. Megens1,3; Alisina Sarabi2; Sabine Winkler2; Delia Projahn2; Christian Weber1,3,4; Rory R. Koenen1,3,*; Philipp von Hundelshausen1,4,* 1Institute

for Cardiovascular Prevention, Ludwig-Maximilians-University of Munich, Munich, Germany; 2Institute for Molecular Cardiovascular Research (IMCAR), Medical Faculty, RWTH Aachen University, Aachen, Germany; 3Cardiovascular Research Institute Maastricht (CARIM), Maastricht University, Maastricht, The Netherlands; 4DZHK (German Centre for Cardiovascular Research), partner site Munich Heart Alliance, Munich, Germany

Summary The chemokine CCL5 recruits monocytes into inflamed tissues by triggering primarily CCR1-mediated arrest on endothelial cells, whereas subsequent spreading is dominated by CCR5. The CCL5-induced arrest can be enhanced by heteromer formation with CXCL4. To identify mechanisms for receptor-specific functions, we employed CCL5 mutants and transfectants expressing receptor chimeras carrying transposed extracellular regions. Mutation of the basic 50s cluster of CCL5, a coordinative site for CCL5 surface presentation, reduced CCR5- but not CCR1-mediated arrest and transmigration. Impaired arrest was restored by exchanging the CCR5-N-terminus for that of CCR1, which supported arrest even without the 50s cluster, whereas mutation of the basic 40s

Correspondence to: Rory R. Koenen Cardiovascular Research Institute Maastricht (CARIM) Maastricht University, Maastricht, The Netherlands Tel.: +31 43 38 83390; Fax: +31 43 3884159 E-mail: [email protected] or Philipp von Hundelshausen Institute for Cardiovascular Prevention Ludwig-Maximilians-University of Munich Munich, Germany Tel.: +49 89 5160 4359, Fax: +49 89 5160 4352 E-mail: [email protected]

cluster essential for proteoglycan binding of CCL5 could not be rescued. The enhancement of CCL5-induced arrest by CXCL4 was mediated by CCR1 requiring its third extracellular loop. The domain exchanges did not affect formation and co-localisation of receptor dimers, indicating a sensing role of the third extracellular loop for hetero-oligomers in an arrest microenvironment. Our data identify confined targetable regions of CCR1 specialised to facilitate CCL5-induced arrest and enhanced responsiveness to the CXCL4-CCL5 heteromer.

Keywords Arrest, chemotaxis, inflammation, chemokine receptor chimera, RANTES

Received: May 24, 2013 Accepted after minor revision: June 30, 2013 Prepublished online: August 8, 2013 doi:10.1160/TH13-05-0420 Thromb Haemost 2013; 110: 795–806

* These authors share senior authorship. Note: The review process for this manuscript was fully handled by G. Y. H. Lip, Editor in Chief.

Introduction The coordinated movement of cells is a fundamental requirement for multicellular organisms during their development and for the correct function of their immune system. Chemoattractant cytokines, termed chemokines and their corresponding receptors play a major role in the cellular guidance-system of the body. Beside their physiological relevance, the pathophysiological aspects of chemokines are of great interest. In a vast number of studies, chemokines were shown to be causally linked to many diseases, including atherosclerosis, asthma, rheumatoid arthritis, and many more (1–3). The chemokine system consists of more than 20 receptors and over 50 cognate ligands identified to date (4) permitting an enormous combinatorial diversity, since a given chemo© Schattauer 2013

kine can bind several receptors and vice versa. CCL5 (RANTES) binds with high affinity to CCR1 and CCR5, and CCR5 binds in addition to CCL5 seven other chemokines (5) which not necessarily bind to CCR1 such as CCL4 (MIP-1β). An often used adjective for this phenomenon is “redundant”. However, this is a somewhat misleading description because the binding of different chemokines to one receptor must not necessarily result in the same response (6). Furthermore, functional specialisation and locally restricted presentation of chemokines and their receptors serve to add a degree of specificity to the chemokine system, rather than redundancy (6, 7). This is exemplified by CCL5 and its receptors CCR1 and CCR5. Our previous studies have suggested specialised roles for CCR1 and CCR5 in leukocyte recruitment (8, 9). CCR1 was found to be predominantly required for the initial adhesion of Thrombosis and Haemostasis 110.4/2013

Downloaded from www.thrombosis-online.com on 2014-02-10 | ID: 1000364783 | IP: 137.120.195.152 For personal or educational use only. No other uses without permission. All rights reserved.

796

Kramp et al. Role of the extracellular domains of CCR1 and CCR5

human monocytes and T cells to endothelial cells, whereas CCR5 appeared to mediate in the first line the spreading on endothelial cells preceding transendothelial migration (9). The propensity of CCL5 to form higher-order oligomers has been shown to be particularly important for triggering flow-resistant cell arrest via CCR1, whereas transendothelial migration mediated by CCR5 did not require the oligomerisation of CCL5 (8). The structural requirements for the oligomerisation of CCL5 and also for its binding to heparin-like glycosaminoglycans (GAGs) have been extensively characterised (10–12). Mutation of acidic residues

E26 and E66 into neutral amino acids reduced the oligomerisation to the level of tetramers and dimers, respectively (10). Further, CCL5 contains two basic amino acid clusters located in the 40s and 50s loop regions that were involved in binding to GAGs and cell surfaces (12–14). Neutralisation of the 40s cluster resulted in 44AANA47-CCL5 (CCL5_40s), a variant with strongly reduced GAG-binding and an 80-fold reduction in affinity selectively for CCR1 but a normal binding to CCR5 (12). Neutralisation of the 50s cluster (55KKWVR59) in CCL5 (CCL5_50s) showed an almost unaltered binding to heparin, CCR1 and CCR5 (12). However, the bind-

A

B

C

D

Figure 1: Design and expression of the chimeric chemokine receptors. A) Schematic presentation of an alignment of the CCR1 amino acid sequence with CCR5. Topographic blot based on the CCR1 amino acid sequence; blue circles denote identical residues in CCR5 through CCR1 with annotated disulfide bonds (dotted lines). B) Sketches illustrating the mutations of the CCL5 variants. C) Schematic diagram of the different CCR1–

CCR5 chimeras. The CCR5 part of the various chimeras is marked in blue and the CCR1 domains in gray. D) Expression of CCR5, CCR1 and the related mutants on L1.2 pre-B lymphoma cells was analysed by flow cytometry. The black lines represent the specific mAb against CCR1 or CCR5 and the gray lines illustrate the isotype control.

Thrombosis and Haemostasis 110.4/2013 Downloaded from www.thrombosis-online.com on 2014-02-10 | ID: 1000364783 | IP: 137.120.195.152 For personal or educational use only. No other uses without permission. All rights reserved.

© Schattauer 2013


ing of CCL5_50s to endothelial cells and tissue specimen was abolished (13), and in vivo peritoneal leukocyte recruitment was defective when CCL5_50s was injected, implying a crucial role in vivo for CCL5 presentation by cell surface molecules distinct from heparin proteoglycans (14). In fact, by forming GAG-binding-deficient heterodimers with CCL5, CCL5_40s acts as a dominant-negative inhibitor in vivo (15), and exogenously administered CCL5_40s effectively blocked the pathophysiologic activities of CCL5 in mice, e.g. during the progression of atherosclerosis (16). Interestingly, heteromerisation of CCL5 also takes place with endogenous ligands under physiologic conditions. Notably, through binding to CCL5, the platelet-derived chemokine CXCL4 (platelet factor 4) was shown to enhance CCL5-induced monocyte recruitment under flow conditions (17). Using a synthetic peptide interfering with this binding, we could demonstrate that the synergistic interaction of CCL5 and CXCL4 was involved in monocyte recruitment during the progression of atherosclerotic lesions (18). Although the above data indicate that oligomerisation and heteromerisation of CCL5 are crucial for its in vivo function, the molecular mechanisms underlying its actions at the receptor level remain poorly defined. Similar applies for the functional specialisation of CCR1 and CCR5. To establish which domains are important for the functional interaction between these receptors and CCL5, we exchanged the four extracellular regions of CCR5 with those from CCR1 generating stably transfected chimeric CCR5 variants in L1.2 and HEK293 cells and characterised their recruitment by CCL5 variants or adding CXCL4 to discriminate important functional interfaces of CCL5 and its receptors.

Material and methods Chemokine receptor cDNA constructs The pcDNA3.1-based plasmids containing the cDNAs for human CCR1 (hCCR1) (accession P32246) and human CCR5 (hCCR5)

Table 1: Nomenclature of chemokine receptor constructs.

© Schattauer 2013

(accession P51681) were purchased from the Missouri S&T cDNA Resource Center (Rolla, MO, USA). The CCR1 sequence was aligned against the CCR5 sequence (19) using the BLOSUM62 matrix. Sequences were aligned using CLUSTAL-W (20) (▶ Figure 1 A). The length of the extracellular loops (ECL) was determined according to the UniProtKB/Swiss-Prot database, considering a minimum length of 18 residues in the core of the lipid double membrane and criteria from sequence analysis. Chimeric constructs of CCR5 containing the extracellular domains of CCR1 and “reverse” receptor constructs of CCR1 containing extracellular domains of CCR5 (▶ Figure 1 B) were designed in-house and purchased from GenScript USA Inc. (Piscataway, NJ, USA), in pUC57. These chimeric receptors were subsequently cloned into the pcDNA3.1 or the pcDNA4A plasmid (Life Technologies, Darmstadt, Germany). In the same manner CCR1, CCR5 and CCR1–5N1 or CCR1–5E3 containing a synthetic C-terminally Myc-tag or a hemagglutinin (HA) tag were constructed. The nomenclature and detailed of the mutant constructs is summarised in ▶ Table 1 and in Suppl. Figure 1 (available online at www.throm bosis-online.com.

Cell culture and transfection Murine pre-B lymphoma L1.2 cells and human embryonic kidney cells (HEK293, ATCC CRL 1573, Manassas, VA, USA), were stably transfected with liposomes (Fugene®, Roche, Basel, Switzerland) after cells reached 50% confluence, to express hCCR1, hCCR5 and the CCR1/CCR5 chimeras described in ▶ Table 1 and ▶ Figure 1 A, followed by cell selection according to the highest expression. L1.2 cells were transfected by nucleofection using the Amaxa Nucleofector™ kit V and a Amaxa II Nucleofector™ System (Lonza AG, Visp, Switzerland). Briefly, 2 x 106 cells were resuspended in 100 μl of nucleofection solution, and after addition of 2 μg of the indicated expression vector, the cells were immediately pulsed

Construct name

Description

Exchanged amino acids

CCR5_1N1

hCCR5 containing the N-terminus of hCCR1

CCR5 M1R30 to CCR1 M1Q35

CCR5_1E1

hCCR5 containing the first ECL of hCCR1

CCR5 H92Q107 to CCR1 D92K107

CCR5_1E2

hCCR5 containing the second ECL of hCCR1

CCR5 T172I203 to CCR1 S172L203

CCR5_1E3

hCCR5 containing the third ECL of hCCR1

CCR5 Q266Q282 to CCR1 Q266L282

CCR1_5N1

hCCR1 containing the N-terminus of hCCR5

CCR1 M1A34 to CCR5 M1A30

CCR1_5E3

hCCR5 containing the third ECL of hCCR1

CCR1 Q266Q282 to CCR5 Q266Q282

CCR5 and CCR5_1E3 _HA) CCR1 and CCR1_5E3 _Myc)

wild-type or chimeric receptors containing a C-terminal Mycor HA-tag


797

798


using the U-15 program. Selection was carried out using neomycin (600 µg/ml). Surface expression levels were analysed by flow cytometry (FACS) (FACSCantoII or the FACSCalibur, BD Biosciences, San Jose, CA, USA). Cells were stained with mouse α-hCCR1-phycoerythrin (α-hCCR1-PE, clone 53504) or mouse α-hCCR5-allophycocyanin (α-hCCR5-APC, clone CTC5) mAB or treated with the respective isotype-controls (mouse IgG2B-PE and mouse IgG1-APC). All antibodies were purchased from R&D Systems (Wiesbaden, Germany).

Chemotaxis The Costar Transwell™ system (5.0-µm pore size) (Corning Glass, Corning, NY, USA) was used to quantify the migration of HEK293 cells stably expressing CCR1, CCR5 or CCR1/CCR5 chimeras towards CCL5 and the CCL5_40s and CCL5_50s mutants (100 ng/ ml). After 4 hours (h), the inserts were removed and the number of migrated cells in the lower well was counted for 1 minute (min) using the high throughput loader on a Becton Dickinson FACSCantoII (BD Biosciences) with the gates set to acquire the cells of interest.

Adhesion assays Adhesion of stably transfected L1.2 cells, labelled with calcein (Life technologies) to a mouse tumor necrosis factor (TNF) α-activated (10 ng/ml, 4 h) monolayer of SV40-immortalised mouse endothelial cells (SVEC, ATCC CRL-2181) was analysed in a parallel wall chamber under laminar flow conditions (0.1 ml/min, 1.5 dynes/cm2). The L1.2 transfectants were activated with H2O (control), CCL5 and mutants (500 ng/ml) or/and CXCL4 (500 ng/ml) at 37°C prior to perfusion. During perfusion, stably adherent cells were counted in at least 10 random microscopic fields using microscopic video imaging (21). Adhesion of human acute monocytic leukemia cell line (THP-1, ATCC TIB-202) to Human Aortic Endothelial Cells (HAoEC) activated with human TNF-α (10 ng/ ml, 4 h) directly stimulated with CCL5 (500 ng/ml) or/and CXCL4 (500 ng/ml) was determined using the same method. THP cells were pretreated with pertussis toxin (PTX, 100 ng/ml, 30 min, 37°C, Sigma-Aldrich, St. Louis, MO, USA) or left untreated. Stable cell arrest was assessed in the presence or absence of either the CCR5 antagonist DAPTA (10 nM) or J113863 antagonising CCR1/CCR3 (10 nM) or SB297006 a selective CCR3 (80 nM) antagonist (Tocris Bioscience Bristol, UK).

Mutants of CCL5 All CCL5 mutant constructs were produced by recombinant methods in Escherichia coli as previously described (12, 18) except CCL5_NmeT7, which is an obligate monomeric synthetic CCL5 variant (22). The CCL5_E66A mutant is a tetrameric mutant (10, 22) and the CCL5_40s and CCL5_50s mutants contain alanine substitutions in their basic amino acid clusters 44RKNR47 and

55KKWVR59

both involved in the binding of CCL5 to glycosaminoglycans (12, 13).

Immunofluorescence The wild-type HEK293 cells were seeded on round glass cover slides coated with polylysine the day before transfection with pcDNA3.1-CCR1-Myc, pcDNA3.1-CCR5-HA or pcDNA4– CCR1_5E3-Myc or pcDNA3.1-CCR5_1E3-HA. Liposomal transfection (Fugene®, Roche) was carried out after cells reached 60–70% confluence, subsequently the cells were incubated for 48 h at 37°C and 5% CO2. All following steps were carried out at room temperature. Cells were fixed in 2% (w/v) paraformaldehyde-PBS for 15 min and then permeabilised in 0.2% (v/v) saponin diluted in PBS for 25 min. After blockage with 1% bovine serum albumin (BSA) (v/v) in phosphatebuffered saline (PBS), the cells were sequentially stained with 20 µg/ml anti-Myc-fluorescein isothiocyanate (FITC, Life Technologies, Darmstadt, Germany) and anti-HA Alexa Fluor® 594-conjugated mAb (Life Technologies) at a concentration of 50 µg/ml. Finally the cells were counterstained with DAPI (Life Technologies). The cover slides were than mounted on a microscope slide. The cell proteins were visualised using a two-photon microscope (Leica SP5II MP, Mannheim, Germany) set to confocal mode as previously described (23).

Co-immunoprecipitation HEK293 cells were transiently transfected with pcDNA4.1CCR1-Myc, pcDNA3.1-CCR5-HA or pcDNA4–CCR1_5E3-Myc or pcDNA3.1-CCR5_1E3-HA (vide supra). For pull-down of the CCR1/CCR5 receptor complexes the transfected as well as untransfected HEK cells were lysed in non-denaturing lysis buffer 1% (w/v) Triton X-100, 1% (w/v) CHAPS (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate) (Calbiochem), 50 mM Tris-HCl (pH 7.4), 15mM EGTA, 100 mM NaCl (Carl Roth), supplemented with a proteinase inhibitor cocktail (Roche, Mannheim, Germany) to a final concentration of 5 x 106 cells/ml. Supernatants were incubated with 10 µg of the appropriate antibodies, aHA-Tag (C29F4) or aMyc-Tag (71D10) (Cell signalling, Danvers, MA, USA), respectively, bound to Dynabeads® protein G beads (Life technologies). Proteins were analysed by western blot using antibodies to the Myc-Tag and the HA-Tag directly conjugated to HRP.

Statistical analysis Data are expressed as mean ± standard error of the mean. The number of independent experiments was 5 or, when otherwise, was indicated in each figure legend. The data were tested for a Gaussian distribution using the Kolmogorov-Smirnov test. A nonparametric Kruskal-Wallis test was used when the data were nonGaussian distributed. Calculations were performed with GraphPad Prism version 5.04 for Windows (GraphPad Software, La Jolla, CA, USA; www.graphpad.com). Comparisons of all pairs were performed using Newman-Keuls or Dunn’s multiple comparison tests, as appropriate. A p-value below 0.05 was considered as statistically significant.


© Schattauer 2013


Results Design and characterisation of the chemokine receptor chimeras Our previous reports have revealed a functional specialisation of CCR1 and CCR5 in the recruitment of monocytes and T cells, whereby CCR1 was more important in promoting stable arrest on endothelial cells and CCR5 primarily mediated subsequent spreading and transendothelial migration (8, 9). The three-dimensional structure of a chemokine binding to its receptor has not yet been solved, but it is believed that the extracellular loops predominantly interact with larger ligands such as chemokines or heteromers as opposed to small molecule antagonists that bind deep in the ligand pocket located between the extracellular parts of the transmembrane domains (24, 25). In contrast to the highly conserved transmembrane regions of CCR1 and CCR5, sequence analysis demonstrates that the extracellular regions are remarkably heterogeneous (▶ Figure 1 A). In order to identify the role of these extracellular regions for leukocyte arrest and migration, we generated chimeric constructs by exchanging the respective extracellu-

Figure 2: Chemotaxis of receptor-transfectants towards CCL5 variants. Migratory response of HEK293 cells transfected with CCR1 (A, Kruskal-Wallis, p = 0.0042), CCR5 (B, ANOVA, p < 0.0001), CCR5_1N1 (C, ANOVA, p < 0.0001), and CCR5_1E1–3 (D, Kruskal-Wallis, p < 0.0001, E, p = 0.0169, F, p = 0.0027) towards CCL5, CCL5_50s, CCL5_40s (each 100 ng/ml) or negative control expressed as chemotactic index (CI = ratio to the control). (*P